Ning Cao, Staff Engineer, Nano-fabrication lab, UCSB Operation Manual of J.A. Woollam Ellipsometer 1) Press both the Lamp power and lamp Ignition buttons to turn them on (The buttons are located on the front panel of the bottom control box). 2) Put your film sample onto the center of the sample stage (there are two tiny vacuum holes underneath your sample). 3) Switch the vacuum on (at the low left side of the tool). 4) Open CompleteEASE software (if it is not opening). 5) Click on the Measurement tab at the top left of screen, then, click on the arrow-down sign, the recipe menu showing up, then, choose Standard. 6) Click on Measure tab, then, a Panel shows up. Make sure: a) Acq. Time is 4.00; b) High Accuracy mode is ticked; c) Angle Scan: 55 To 75 By 10.00; d) Sample tilt Alignment: Manual; e) Sample Height Alignment: Automatic-Quick; f) Sample Thickness: type the thickness of your sample substrate, e.g., 0.50 mm; g) Align At First Angle is not ticked; h) Alignment Angle: 65; i) Do not Return To Sample Load Position is ticked; j) Finally, click on OK tab at the bottom of the Panel. 7) Sample tilt Alignment starts: a) if there is some reflected light signal entering the detector aperture hole, a blue-color section appears on the screen, just adjust the two knobs located at the two sides of the sample stage to center the red cross sign, then, click on Cancel Alignment tab at the top left of screen (or just hit the ESC key on the keyboard) to end the alignment (sometimes, one needs to repeat the alignment several times before the measurement starts). b) In some situation, there is no blue-color section appearing on the screen and the red-cross sign jumps all over the screen: this means the detector does not receive any signal reflected from the sample. In such case, what you need to do is looking for a circled white-light spot (weak, but, you can see it) outside the aperture hole, then, adjust the two knobs to make the spot moving toward the aperture hole, as soon as there is some signal light entering the aperture hole, the blue-color section appears on the screen, keep adjusting these two knobs to make the red cross sign to be centered, while the blue-color section becomes larger and larger. 8) After the alignment, the measurement starts at the multiple angles. Wait until it is finished: Psi and Delta spectral curves showing up in the Graph section. 9) Click on Analysis tab at the top left of screen. Then, there are 4 sections appearing in the Analysis screen: Data, Fit, Model, and Graph ones. 10) If you want to save the measurement data, click on the Save tab in the Data section, choose your file directory, type a file name, and save the data into the file (you can click on Open tab in the Data section to open the data file for later-on analyses).
Transcript
Ning Cao, Staff Engineer, Nano-fabrication lab, UCSB
Operation Manual of J.A. Woollam Ellipsometer
1) Press both the Lamp power and lamp Ignition buttons to turn them on (The buttons are located on the front panel of the bottom control box).
2) Put your film sample onto the center of the sample stage (there are two tiny vacuum holes underneath your sample).
3) Switch the vacuum on (at the low left side of the tool).
4) Open CompleteEASE software (if it is not opening).
5) Click on the Measurement tab at the top left of screen, then, click on the arrow-down sign, the recipe menu showing up, then, choose
Standard.
6) Click on Measure tab, then, a Panel shows up. Make sure: a) Acq. Time is 4.00; b) High Accuracy mode is ticked; c) Angle Scan: 55 To 75 By
10.00; d) Sample tilt Alignment: Manual; e) Sample Height Alignment: Automatic-Quick; f) Sample Thickness: type the thickness of your sample
substrate, e.g., 0.50 mm; g) Align At First Angle is not ticked; h) Alignment Angle: 65; i) Do not Return To Sample Load Position is ticked; j)
Finally, click on OK tab at the bottom of the Panel.
7) Sample tilt Alignment starts: a) if there is some reflected light signal entering the detector aperture hole, a blue-color section appears on the
screen, just adjust the two knobs located at the two sides of the sample stage to center the red cross sign, then, click on Cancel Alignment tab at
the top left of screen (or just hit the ESC key on the keyboard) to end the alignment (sometimes, one needs to repeat the alignment several
times before the measurement starts). b) In some situation, there is no blue-color section appearing on the screen and the red-cross sign jumps
all over the screen: this means the detector does not receive any signal reflected from the sample. In such case, what you need to do is looking
for a circled white-light spot (weak, but, you can see it) outside the aperture hole, then, adjust the two knobs to make the spot moving toward
the aperture hole, as soon as there is some signal light entering the aperture hole, the blue-color section appears on the screen, keep adjusting
these two knobs to make the red cross sign to be centered, while the blue-color section becomes larger and larger.
8) After the alignment, the measurement starts at the multiple angles. Wait until it is finished: Psi and Delta spectral curves showing up in the
Graph section.
9) Click on Analysis tab at the top left of screen. Then, there are 4 sections appearing in the Analysis screen: Data, Fit, Model, and Graph ones.
10) If you want to save the measurement data, click on the Save tab in the Data section, choose your file directory, type a file name, and save
the data into the file (you can click on Open tab in the Data section to open the data file for later-on analyses).
Ning Cao, Staff Engineer, Nano-fabrication lab, UCSB
11) Click on Open tab in the Model section to open a model (you can create a new model or copy some existing model and edit it, including to
add or delete layer(s) under Layer Commends category, if necessary. Then, click on Save tab in the Model section to save it with a different
name into your directory). For your reference: 1) using Cauchy model, n ()=A+B/2+C/4, where A, B, C are the Cauchy fitting parameters, for a
dielectric film or a photoresist film; 2) using B-Spline model for a thin absorptive metal film (to get a good result, the thickness of the metal film
needs to be less than 50 nm); 3) using Si_JAW for a Si substrate or a single-crystal Si layer in SOI sample structure; 4) using NTVE_JAW for a very
thin oxide layer (turning the fit operation on for the layer thickness) or a native oxide layer with a fixed thickness of 1 nm.
12) Click on the Fit tab in the Fit section to fit the data with the model chosen. You can set the wavelength range by clicking on Set Ranges tab in
the Fit section: a panel shows up, then, typing the new low or/and high wavelength limit(s) there (you can also select the data of a certain angle
to fit), finally, hitting OK tab. A good fit means a small Mean-Squared-Error (MSE) value.
13) To get the optical constants at a certain wavelength (see Figure 1 below for a reference), before fitting, do the following: click on the red plus
sign beside FIT Options in the Model section; click on the red plus sign beside Include derived Parameters=ON (if it is OFF, just click on OFF sign
to change to ON) at the bottom; choose Type, and type Layer # and Wavelength in which you are interested. You can add as many Derived
Parameters as you want and these derived parameters will show up after the fit.
14) To get the optical constant spectra (see Figure 2 below for a reference), n() and k(), after fitting data curves, do the following: click on
Options tab at the top left of the screen; choose n&k in Optical Constants (or e1&e2); then, click on Analysis tab at the top left of screen; right-
mouse-button click on the Model name of the layer, you are interested in getting the optical constants, in the Model section, a menu showing
up; choose Graph Layer Optical Constants; the n() and k() [or e1()&e2()] spectra show up in the Graph section; right-mouse-button click on
at any point within the Graph area, a menu showing up; choose Copy Data to Clipboard; open Microsoft Excel file; paste the data into a
spreadsheet; save the spreadsheet data into an Excel file.
15) If a film is deposited on a double-side-polished substrate, you need to do a correction to reduce the error caused by the backside-reflection
(see Figure 3 below for a reference): click on the red plus sign beside MODEL Options in the Model section; click on the blue OFF sign of Include
Substrate backside Correction=OFF to turn the correction on (the blue ON sign showing up); click on the Configure Options at the bottom of the
Model section: a frame showing up; choose Include Depolarization Data by clicking on the small square empty box (a tick showing up) in the
Available Options list under Fit Options , then, hit Ok tab; click on the red plus sign beside Fit Options in the Model section; click on the OFF sign
of Include Depolarization Data=OFF to turn the function on (the blue ON sign showing up).
16) Save the data as well as fitting result in a Snapshot file for a later-on analyses: click on the Save Snapshot tab at the top right of screen, the
Save Snapshot frame showing up; choose your file directory, type a file name, and click on the Save tab (you can click on Open Snapshot tab to
open the snapshot file for later-on analyses).
Ning Cao, Staff Engineer, Nano-fabrication lab, UCSB
17) When you done, switch the vacuum off, remove your sample off the stage, and, most importantly, Turn Lamp Power off to save the lamp
lifetime.
18) Let me know if you have some issues (Ning Cao, 805-893-4689 or [email protected]).
Figure 1 How to get the opical constants at the wavelengh(s) you are interested.
Ning Cao, Staff Engineer, Nano-fabrication lab, UCSB
Figure 2 How to get the opical constant spectra [n()&k() or e1()&e2()].
Ning Cao, Staff Engineer, Nano-fabrication lab, UCSB
Figure 3 How to set up Backside Reflection Correction for a film deposited on a double-side polished substrate.
Ning Cao, Staff Engineer, Nano-fabrication lab, UCSB
Examples:
Example A: a low-stress (LS) SiNx film on a one-side polished Si substrate. Fitting the data using Cauchy mode, n () =A+B/2+C/4, where A, B, C
as well as film thickness are the fitting parameters.
Figure 4 Shows the fit result of the LS SiNx ellipsometry data: film thickness=254.25 nm and MSE= 172.055 with the fit range between 190 and
1688 nm.
Ning Cao, Staff Engineer, Nano-fabrication lab, UCSB
m.
The MSE value (172.055) is too large which is due to the fitting at lower wavelengths is poor. This can be checked out by the Depolarization
spectrum [click on Graph Type menu in the middle left of screen, then, Depolarization (see Figure 5 below)]. The depolarization at low
wavelengths is not ideal: there are several peaks, especially, the two big ones below 300 nm.
Ning Cao, Staff Engineer, Nano-fabrication lab, UCSB
Figure 5 Depolarization spectrum of the LS SiNx film.
What you can do to get a better fit with a much smaller MSE value is to reset the fit wavelength range: in this case, to reset the range to
between 300 and 1688 nm (see the result in Figure 6 below), then, fit the data curves again. The new fit result is much better with a MSE value
of 16.376.
Ning Cao, Staff Engineer, Nano-fabrication lab, UCSB
Figure 6 Shows the fit result of the LS SiNx ellipsometry data: film thickness=255.20 nm and MSE= 16.376 with the fit range between 300 and
1688 nm.
Example B: Amorphous Si film, deposited using PECVD#2, on a 500-nm thermal SiO2 (just called SiO2 below)/one-side polished Si substrate: 1)
Taking the measurement of SiO2/Si, fitting the data using Cauchy model, and saving it into a Snapshot file; 2) Depositing a-Si film on the SiO2/Si
Ning Cao, Staff Engineer, Nano-fabrication lab, UCSB
substrate and measuring the Psi and Delta spectra of a-Si/SiO2/Si; 3) opening the Snapshot file of SiO2/Si, fixing all the fitting parameters of the
first SiO2 layer as well as Angle Offset value in the model, and opening the data file of a-Si/SiO2/Si; 4) Adding the second layer with B-Spline
model for a-Si to the model; 5) loading the a-Si Parameterized file from Semiconductor directory into the Starting Mat and turning Use KK Mode
on ; 6) Fitting the a-Si/SiO2/Si data with this modified model to get the result (see Figure 7 below): a-Si film thickness of 31.17 nm (MSE=4.460).
Figure 7 Fitting result of a-Si/SiO2/Si (a-Si film thickness =31.17 nm and MSE=4.640).
Ning Cao, Staff Engineer, Nano-fabrication lab, UCSB
Example C: Cr/Ta/thermal SiO2 (just called SiO2 below) on one-side polished Si substrate: 1) Taking the measurement of SiO2/Si, fitting the data
using Cauchy model (see Figure 8), and saving it into a Snapshot file.
Figure 8 Fitting result of SiO2/Si (film thickness =301.27 nm, A=1.447, B=0.00589, C=-0.00012025, and MSE=8.497).
Ning Cao, Staff Engineer, Nano-fabrication lab, UCSB
2) Depositing a thin Ta film (the target film thickness is 30 nm) onto the SiO2/Si substrate. Then, taking the ellipsometry measurement of
Ta/SiO2/Si and saving the data into a data file. Then, opening the previous Snapshot file of SiO2/Si with the SiO2/Si data curves, and opening the
Ta/SiO2/Si data file. Then, editing the model by adding the second layer with a B-Spline model for the Ta film, loading the Ta file from the metal
directory into Starting Mat and turning Use KK Mode on, fixing the fitting parameters of the first layer (SiO2) as well as the Angle Offset value.
Finally, fitting the Ta/SiO2/Si data curves using this modified model, and saving it into another Snapshot (see Figure 9 below).
Figure 9 Fitting result of Ta/SiO2/Si (Ta film thickness =35.24 nm and MSE=4.726).
Ning Cao, Staff Engineer, Nano-fabrication lab, UCSB
3) Depositing a thin Cr film (the target film thickness is 30 nm) onto the Ta/SiO2/Si substrate. Then, taking the ellipsometry measurement of
Cr/Ta/ SiO2/Si and saving the data into a data file. Then, opening the previous Snapshot file of Ta/SiO2/Si with the Ta/SiO2/Si data curves, and
opening the Cr/Ta/SiO2/Si data file. Then, editing the model by adding the third layer with a B-Spline model for the Cr film, loading the Cr
(Lorentz) file from the metal directory into Starting Mat and turning Use KK Mode on, fixing the fitting parameters of the second layers (Ta).
Finally, fitting the Cr/Ta/SiO2 data curves using this modified model, and saving it into a Snapshot (see Figure 10 below).
Ning Cao, Staff Engineer, Nano-fabrication lab, UCSB
Figure 10 Fitting result of Cr/Ta/SiO2/Si (Cr film thickness =34.85 nm and MSE=2.310).
